In order to complete a survey of a possible developmental pathway leading to defiance and other antisocial behavior in male children from stressed low income families, it is proposed to carry out follow-up assessments at 57 and 72 months with an initial sample of 310 males previously evaluated at 18, 24, and 42 months. The hypotheses are that insecure attachments in the first year of life result from the interactions of demanding infants and unresponsive mothers who, living in limited space and under economic stress, engage in power struggles between 12 and 24 months, when the infants become mobile and more autonomous. As a result, the infants become more noncompliant and the mothers become more manipulative, leading to coercive cycles of interaction between 2 1/2 and 3 1/2 years. It is further hypothesized that between 3 1/2 and 6 years, the coercive cycles of interaction spread to other members of the family, and the parents' discipline becomes inconsistent and ineffective. By age 6, the child is impulsive and restless, and becomes increasingly defiant with parents, aggressive and hostile with siblings and peers. The hypotheses involve tracing developmental events for the entire period from 18 to 72 months and, linking extensive research that has been carried out on attachment up to the preschool period, on the one hand, and on coercive cycles of parent-child interaction in the case of early school-age children on the other. The two latter bodies of research as yet have not been integrated empirically on low income samples despite their greater risk for developing later externalizing behavior problems. Because the follow-up covering 18, 24, and 42 months was considered too much of an undertaking given the time constraints and budgetary restrictions of a FIRST Award, the review committee recommended a separate submission for a later follow-up, provided it was logistically feasible. Now it appears both feasible and desirable to carry out a follow-up at 57 and 72 months since recruitment of the sample is complete, data collection of the first two assessments points is nearly finished, and data from a pilot study and the present investigation indicate results supporting a developmental pathway leading to early disruptive behavior from 12 to 36 months. The present application includes two 57-month home visits and a 72-month laboratory assessment so we can continue to pursue this pathway.